I do not no exactly what observations were available then, but what geophysicists do is to look at fossil indications from millions of years ago in the form of the growth rings in coral, sea shells, and sediment deposition at the bottom of lakes. These seem to form layers that, when dated, yield a different time interval for the lunar high and low tides. I believe that records of ancient eclipses seen in specific parts of the world several thousand years ago, when reproduced by modern forecasts, require a slight change for the faster rotation of the Earth to be made. Today we know that the Moon is moving away from the Earth at about 3.8 centimeters per year from direct observation, and that the rotation of the Earth is also slowing down at a rate that has been constant for at least the last 900 million years.